2018 FIDE Rapid & Blitz World Championships

[Richard Bates]

The then Chairman of the FIDE Qualification Commission promptly decreed that nothing similar would be permitted again.

To be fair, this was done at the behest of Swiss Pairings, who considered that their pairing rules were the only ones acceptable for norm events.

I do intend to renew my input to the new Commission, to the effect that the rules which allow the top seed in this event to as many as five "leg-ups" as and when required are grotesquely biased. We shall see how this is received.

Isn’t the problem the snake down as much as the ladder up? The whole point of FIDE rules is to “knock out” the weaker players at minimal disruption (although as has been noted before it fails particularly in the first round).With that purpose in mind you need to give them the strongest possible opponent. I believe that FIDE rules don’t even require a colour match to take priority. In some ways the current rules restricting the number of floats are already a compromise.

[NickFaulks]

Carlsen finishes the day half off the lead so all to play for on day 3.

With the benefit of the 400 point limit, he has lost only 33 rating points!

[NickFaulks]

The whole point of FIDE rules is to “knock out” the weaker players at minimal disruption

The second seed is likely to be just as effective in this role. My complaint is that it is allocated to the top seed, over and over again.

In any case, in a long event such as this, knocking out weaker players really doesn't need to be a priority, it will happen soon enough in the normal course of events. The real priority should be to create a level playing field for the contenders.

[Chris Rice]

Carlsen finishes the day half off the lead so all to play for on day 3.

Which included another couple of Magnus getting something out of absolutely nothing victories. The joint leaders at the moment on 7½/10 are Nepomniachtchi, Ian (2771), Andreikin, Dmitry (2725), Wang, Hao (2782), Yu, Yangyi (2758), Artemiev, Vladislav (2812), Dubov, Daniil (2723) & Matlakov, Maxim (2690). With 5 rounds to go tomorrow anyone on 6/10 or better has a chance to win this and that's 61 players. Not Gawain though who currently is on 4/10.

Some great play but it was Andreikin's win vs Svidler that really caught the eye as Svidler suddenly realised he was in a mating net.

In the Women's with four rounds to go Wenjun Ju leads with 7/8, half a point ahead of Mariya Muzychuk who she has already played and then there's a bunch on 6. Sue's on 2.5/8.

[Richard Bates]

The whole point of FIDE rules is to “knock out” the weaker players at minimal disruption

The second seed is likely to be just as effective in this role. My complaint is that it is allocated to the top seed, over and over again.

In any case, in a long event such as this, knocking out weaker players really doesn't need to be a priority, it will happen soon enough in the normal course of events. The real priority should be to create a level playing field for the contenders.

I hate the way Fide operate byes and much prefer the CAA approach of “middle floats” (up and down, except in certain circumstances when it is middle down vs top up). But unless you are going to push for this approach to be implemented in FIDE i’m not really sure what you are suggesting. You can’t start designing a pairing system that operates differently according to the nature of the tournament, especially if being done by computers that require a fixed algorithm. And the system already limits the number of floats a player can get. (To one every three rounds). Perhaps the only realistic tweak would be to give a colour match the priority.

And of course the “weak” floating player is a relative term anyway. In any odd numbered group there will be a weakest player. Sometimes it will be the third seed.

[NickFaulks]

I hate the way Fide operate byes and much prefer the CAA approach of “middle floats” (up and down, except in certain circumstances when it is middle down vs top up).

I am of course in agreement with you there. However, for as long as FIDE continues to favour a method which creates the greatest mismatches possible, I would at least include a rule which ( barring exceptional circumstances ) restricts any player to just one of these freebies. In a large tournament this should be easy to satisfy.

[Richard Bates]

I hate the way Fide operate byes and much prefer the CAA approach of “middle floats” (up and down, except in certain circumstances when it is middle down vs top up).

I am of course in agreement with you there. However, for as long as FIDE continues to favour a method which creates the greatest mismatches possible, I would at least include a rule which ( barring exceptional circumstances ) restricts any player to just one of these freebies. In a large tournament this should be easy to satisfy.

But an up float isn’t always a “freebie”. On many occasions upfloating can give a significantly harder opponent than would be available if paired within the score group. I return to the point that the system is designed to float people down to lose. Place too many restrictions on upfloaters and pretty soon floating down becomes a big advantage.

Maybe you are envisaging some sort of system which tries to look at the strength of the opponent floating down before deciding who to float up? Not sure how you put it in a computer algorithm though.

[Chris Rice]

Carlsen finishes the day half off the lead so all to play for on day 3.

Imagine how well he might do if he wasn't "tired and bored"......

Nigel Short: It is the sign of a truly great player when he can play like total crap (by his lofty standards) and still be only half a point off the lead, with five games remaining.

[NickFaulks]

Maybe you are envisaging some sort of system which tries to look at the strength of the opponent floating down before deciding who to float up?

No, I am envisaging precisely what I said I was envisaging.

In a large event, there will be sufficient high rated players in each group who could be given a full point as reward for teaching a low rated overperformer a lesson. It does not have to be the same high rated player every time, or even more than once per tournament.

In the current Rapid event, after ten rounds we have reached the point where it probably doesn't matter any more. In the Blitz, where Carlsen will also enjoy the privileges of top seed, it may take a bit longer.

[Richard Bates]

Maybe you are envisaging some sort of system which tries to look at the strength of the opponent floating down before deciding who to float up?

No, I am envisaging precisely what I said I was envisaging.

In a large event, there will be sufficient high rated players in each group who could be given a full point as reward for teaching a low rated overperformer a lesson. It does not have to be the same high rated player every time, or even more than once per tournament.

In the current Rapid event, after ten rounds we have reached the point where it probably doesn't matter any more. In the Blitz, where Carlsen will also enjoy the privileges of top seed, it may take a bit longer.

Well I don’t really understand, but there we are. You appear to be assuming that players floating down are by definition a low rated overperformer. This is obviously not always the case. It may simply be the lowest rated of an odd number of players who have performed to expectations. And I think it is perverse to say that a player cannot be floated to play Magnus, for no other reason than he mucked up a game 13 rounds earlier in the tournament.

At the end of the day the reason Magnus has had a lot of easy opponents is not because he has been lucky enough to float up a lot. It is simply because he has been so far off the pace that any Swiss pairing system will give him “easy” opponents in such circumstances.

That what Swiss gambits are all about!

[David Sedgwick]

The then Chairman of the FIDE Qualification Commission promptly decreed that nothing similar would be permitted again.

To be fair, this was done at the behest of Swiss Pairings, who considered that their pairing rules were the only ones acceptable for norm events.

I do intend to renew my input to the new Commission, to the effect that the rules which allow the top seed in this event to as many as five "leg-ups" as and when required are grotesquely biased. We shall see how this is received.

Here is an example of the attitude which you used to take to those who follow the Regulations of the Systems of Pairings and Programs Commission:

It seemed to work quite nicely at Hastings last year, but they limited it to 3 rounds. This year where they extended to five rounds (apparently a FIDE requirement) it was a bit of a disaster from a tournament perspective because it was creating “silly” pairings late in the tournament near to the top of the draw when it was therefore having a potential impact on prizes. It didn’t obviously benefit norm seekers either when they could put together a string of good results in the first 5 rounds only to find themselves paired against very low rated players in rounds 6-7. The merits of all accelerated pairing systems inevitably depend to a great extent on the specific ratings profile of the tournament itself.

This year where they extended to five rounds (apparently a FIDE requirement) …

What??? If tournament organisers want to experiment with something strange they have the right to do so, provided that it does not demonstrably favour some players, but they should take responsibility themselves and not invent "FIDE requirements" to hide behind.

A good first step to persuading the SPPC to reform the FIDE Pairing Rules would be for the QC to make their use entirely voluntary.

[NickFaulks]

Well I don’t really understand, but there we are.
.......
That what Swiss gambits are all about!

It has dawned on me that you may not have looked at the details of this case, or perhaps in general..

In round 2 Carlsen was upfloated to an opponent rated 2304, The #2 player in his group had an opponent rated 2485.

In round 5 he upfloated to an opponent rated 2281. The #2 player in his group had an opponent rated 2547.

In round 8 he upfloated to an opponent rated 2502. The #2 player in his group had an opponent rated 2584.

These results are typical of the early and middle rounds of most big tournaments - I have been checking them for years. This is a supercharged version of the Swiss Gambit which works only if you are the top seed in your group ( subject to colour ). Being second seed is no use at all.

[Chris Rice]

Peter Svidler: "Judging by the amount of love my moment of admiration for Dmitry's combination is getting, I should be getting mated, spectacularly, in the center of the board more often. Seriously though - that final position is fabulous, have a look if you haven't yet."

[Brian Towers]

Maybe you are envisaging some sort of system which tries to look at the strength of the opponent floating down before deciding who to float up?

No, I am envisaging precisely what I said I was envisaging.

In a large event, there will be sufficient high rated players in each group who could be given a full point as reward for teaching a low rated overperformer a lesson. It does not have to be the same high rated player every time, or even more than once per tournament.

In the current Rapid event, after ten rounds we have reached the point where it probably doesn't matter any more. In the Blitz, where Carlsen will also enjoy the privileges of top seed, it may take a bit longer.

Well I don’t really understand, but there we are. You appear to be assuming that players floating down are by definition a low rated overperformer. This is obviously not always the case. It may simply be the lowest rated of an odd number of players who have performed to expectations. And I think it is perverse to say that a player cannot be floated to play Magnus, for no other reason than he mucked up a game 13 rounds earlier in the tournament.

At the end of the day the reason Magnus has had a lot of easy opponents is not because he has been lucky enough to float up a lot. It is simply because he has been so far off the pace that any Swiss pairing system will give him “easy” opponents in such circumstances.

That what Swiss gambits are all about!

I have to admit I fail to understand why so many pairs of knickers are being unnecessarily twisted about a tournament where so many of the ratings are random.

Take, for instance, Carlsen's 5th round opponent, Andrew Tang. With a rapid rating of 2281 he was second lowest rated in the 2.5 group. Lowest was Qi B Chen with 2258. Tang was lowest rated in the matching colour group. Next lowest was Vokhidov, 2304. Still, a shockingly weak player. What a gift it must have been for Carlsen. Oh, but wait! Tang's standard rating is more than 200 points higher at 2487 and his blitz rating is 2501. And how is he doing overall (currently after 12 rounds)? Well, he is on 7.5, just 0.5 behind Carlsen on 8. According to Swiss Manager he is in line for a 118 point rise in elo. Perhaps it wasn't such a gift after all? Much ado about nothing?

[Brian Towers]

Round 13. Two players on 9, who have played each other previously, and 8 players half a point behind on 8.5. While below them carnage reins, cue an outbreak of peace and love (perhaps fear?) on those boards as every single game involving a player on 8.5 or 9 ends in a draw!